Table 2 Comparative performance of the PCE and caIRS across validation cohorts

From: Polygenic risk scores improve CAD risk prediction in individuals at borderline and intermediate clinical risk

Model

Cohort

N (cases)

NRI (95% CI)

Sensitivity

Specificity

PPV

NPV

C-index (95% CI)

PCE

ARIC

11008 (774)

–

28.7

92

21.3

94.5

0.758 (0.743–0.773)

caIRS

9.51 (6.44–12.48)

39.7

90.5

24

95.2

0.783 (0.768–0.798)

PCE

MESA

4162 (240)

–

39.6

81.7

11.7

95.7

0.717 (0.689–0.746)

caIRS

5.86 (1.16–11.32)

42.9

84.3

14.3

96

0.732 (0.704–0.761)

PCE

PMBB

14182 (1158)

–

29.4

91.3

23.1

93.6

0.732 (0.719–0.746)

caIRS

9.08 (6.63–11.59)

39

90.8

27.4

94.4

0.761 (0.747–0.774)

PCE

UKB

120590 (3050)

–

25.5

93.8

9.6

98

0.767 (0.759–0.774)

caIRS

9.84 (8.32–11.47)

37.6

91.5

10.3

98.3

0.797 (0.79–0.804)

  1. The Sensitivity, Specificity, Positive Predictive Value (PPV), Negative Predictive Value (NPV), and Net Reclassification Improvement (NRI) values correspond to the 20% risk classification threshold.
Back to article page